The vast majority of all trauma-related amputations in the United States involve the upper limbs. Approximately half of those individuals who receive an upper extremity myoelectric prosthesis eventually abandon use of the system, primarily because of their limited functionality. Thus, there continues to be a need for a significant improvement in prosthetic control strategies. The objective of this bioengineering research program is to develop and clinically evaluate a prototype prosthetic control system that uses imaging to sense residual muscle activity, rather than electromyography. This novel approach can better distinguish between different functional compartments in the forearm muscles, and provide robust control signals that are proportional to muscle activity. This improved sensing strategy has the potential to significantly improve functionality of upper extremity prostheses, and provide dexterous intuitive control that is a significant improvement over current state of the art noninvasive control methods. This interdisciplinary project brings together investigators at George Mason University, commercial partners at Infinite Biomedical Technologies and clinicians at MedStar National Rehabilitation Hospital and Hanger Clinic. Specific Aim 1: To develop and test a compact research-grade sonomyographic prosthetic system We will develop and evaluate a compact low-power embedded system for sonomyography. We will optimize and implement algorithms for real-time classification and control with multiple degrees of freedom (DOF). We will then integrate ultrasound imaging transducers within test prosthetic sockets for testing on individuals with transradial limb loss in a laboratory setting. We will complete system integration and testing and evaluate the sonomyographic signal quality with changes in arm position and socket loading. Specific Aim 2: To evaluate performance of sonomyographic control compared to myoelectric control We will compare the performance of SMG vs myoelectric direct control with mode switching in myoelectric- nave subjects with transradial amputation. Assessment will be performed using a virtual reality Fitts? law task as well as clinical outcome measures using a terminal device. The primary outcome measure will be the SHAP and secondary outcome measure will be the Clothespin Relocation Task. We will assess intuitiveness of control using gaze tracking, and also study quality of movement. We will also compare the performance of SMG vs myoelectric pattern recognition with proportional control in subjects who have been trained on a commercial PR system using the same outcome measures. The successful completion of this project will lead to the first in human evaluation of an integrated prototype that uses low-power portable imaging sensors and real-time image analysis to sense residual muscle activity for prosthetic control. In the long term, we anticipate that the improvements in functionality and intuitiveness of control will increase acceptance by amputees.